Introduction

Valorant is a free-to-play first-person hero shooter developed and published by Riot Games, for Microsoft Windows. First teased under the codename Project A in October 2019, the game began a closed beta period with limited access on April 7, 2020, followed by an official release on June 2, 2020.

The weapons dataset is based on their first ever major tournament, Stage 2: "Masters" of the VCT (Valorant Champions Tour) 2021 which took place between 24th May and 30th May in Iceland.

And the new dataset is taken from the game directly (patch 4.04)

from IPython.display import HTML HTML('''

sns.kdeplot(data=df_diff['Price'], shade=True);

Types of weapons

from plotly.offline import init_notebook_mode, iplot
init_notebook_mode(connected=True)  

weapon_type = dict(Counter(df_new['Weapon Type']))
weapon_type = {'Weapon Type': list(weapon_type.keys()), 'count': list(weapon_type.values())}

fig_weapon = px.pie(weapon_type, values = 'count', names = 'Weapon Type', title = 'Weapon Type Distribution', hole = .5, )
fig_weapon.show()

headshot_dict = return_sorted('HDMG_0')
bodyshot_dict = return_sorted('BDMG_0')



fig_headshot = px.bar(headshot_dict, x = 'weapon', y = 'values', title = 'Weapon Headshot Distribution')
fig_headshot.show()

fig_headshot = px.bar(bodyshot_dict, x = 'weapon', y = 'values', title = 'Weapon Bodyshot Distribution')
fig_headshot.show()

Models :

df_new.describe()

	Price	Fire Rate	Magazine Capacity	Spread ADS	Spread HIP	HDMG_0	BDMG_0	LDMG_0	HDMG_1	BDMG_1	LDMG_1	HDMG_2	BDMG_2	LDMG_2
count	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000	17.000000
mean	1629.411765	7.697647	21.235294	0.779824	1.173529	112.529412	43.352941	35.764706	107.647059	41.352941	34.176471	101.588235	39.294118	32.764706
std	1261.578726	4.913525	23.823647	1.114471	1.443749	65.352044	34.622863	27.673384	67.755942	35.862482	30.406075	70.336743	36.706206	30.918298
min	0.000000	0.600000	2.000000	0.000000	0.100000	24.000000	12.000000	10.000000	20.000000	8.000000	5.000000	9.000000	3.000000	2.000000
25%	800.000000	3.500000	6.000000	0.157000	0.300000	72.000000	26.000000	22.000000	63.000000	22.000000	18.000000	63.000000	22.000000	18.000000
50%	1600.000000	6.750000	13.000000	0.300000	0.450000	95.000000	30.000000	25.000000	88.000000	30.000000	25.000000	77.000000	28.000000	23.000000
75%	2250.000000	12.000000	25.000000	0.790000	1.000000	159.000000	40.000000	34.000000	160.000000	40.000000	34.000000	145.000000	40.000000	34.000000
max	4700.000000	16.000000	100.000000	4.000000	5.000000	255.000000	150.000000	120.000000	255.000000	150.000000	127.000000	255.000000	150.000000	127.000000

df_new=df_new.set_index("Name")

df_anal=df_new.iloc[:, :10]
cols = df_anal.columns.tolist()
cols = [cols[0]]+cols[2:]+[cols[1]]
df_anal = df_anal[cols] 
df_anal

	Weapon Type	Fire Rate	Wall Penetration	Magazine Capacity	Spread ADS	Spread HIP	HDMG_0	BDMG_0	LDMG_0	Price
Name
Classic	Sidearm	6.75	Low	12	0.400	0.40	78	26	22	0
Shorty	Sidearm	3.33	Low	2	4.000	4.00	24	12	10	150
Frenzy	Sidearm	13.00	Low	13	0.450	0.45	78	26	22	450
Ghost	Sidearm	6.75	Medium	15	0.300	0.30	105	30	25	500
Sheriff	Sidearm	4.00	High	6	0.250	0.25	159	55	46	800
Stinger	SMG	16.00	Low	20	0.500	0.65	67	27	22	950
Spectre	SMG	13.33	Medium	30	0.250	0.40	78	26	22	1600
Bulldog	Rifle	10.00	Medium	24	0.300	0.30	115	35	29	2050
Guardian	Rifle	5.25	Medium	12	0.000	0.10	195	65	48	2250
Phantom	Rifle	11.00	Medium	30	0.110	0.20	156	39	33	2900
Vandal	Rifle	9.75	Medium	25	0.157	0.25	160	40	34	2900
Marshall	Sniper	1.50	Medium	5	0.000	1.00	202	101	85	950
Operator	Sniper	0.60	High	5	0.000	5.00	255	150	120	4700
Bucky	Shotgun	1.10	Low	5	2.600	2.60	40	20	19	850
Judge	Shotgun	3.50	Medium	7	2.250	2.25	34	17	14	1850
Ares	Heavy	13.00	High	50	0.900	1.00	72	30	25	1600
Odin	Heavy	12.00	High	100	0.790	0.80	95	38	32	3200

# Correlation matrix
# from https://www.kaggle.com/kerneler/starter-valorant-weapon-stats-f856dcf8-1
def plotCorrelationMatrix(df, graphWidth):

    df = df.dropna('columns') # drop columns with NaN
    df = df[[col for col in df if df[col].nunique() > 1]] # keep columns where there are more than 1 unique values
    if df.shape[1] < 2:
        print(f'No correlation plots shown: The number of non-NaN or constant columns ({df.shape[1]}) is less than 2')
        return
    corr = df.corr()
    plt.figure(num=None, figsize=(graphWidth, graphWidth), dpi=80, facecolor='w', edgecolor='k')
    corrMat = plt.matshow(corr, fignum = 1)
    plt.xticks(range(len(corr.columns)), corr.columns, rotation=90)
    plt.yticks(range(len(corr.columns)), corr.columns)
    plt.gca().xaxis.tick_bottom()
    plt.colorbar(corrMat)
    plt.title(f'Correlation Matrix for the weapons dataframe', fontsize=15)
    plt.show()

plotCorrelationMatrix(df_new, 8)

C:\Users\anass\AppData\Local\Temp/ipykernel_82712/3485586321.py:5: FutureWarning:

In a future version of pandas all arguments of DataFrame.dropna will be keyword-only.

import warnings
warnings.filterwarnings('ignore')
sns.pairplot(df_anal);

sns.pairplot(df_anal,hue='Weapon Type');
#,hue='Gender'

What we can see here us that the damages (Head, body and leg ) are correlated between each other. The price is kind of correlated with the damage and the magazine capacity (with some outliers). And the fire rate is kind of independent from the rest.

First we try with categorical variables, then without categorical variables.

The dataset set is small so there isn't much to draw from it but we are just playing

def get_X_y(columns):
    X = df_anal[columns] #
    Y = df_anal['Price']
    X = pd.get_dummies(data=X)#, drop_first=True)
    return X,Y

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression
import statsmodels.api as sm
from sklearn.metrics import mean_squared_error as msr
from sklearn.preprocessing import MinMaxScaler

def model_analysis(X,Y,loga=False,drop_first=False,preprocessing = True):
    if loga:
        # We shouldnt have any negative values so we might try the Logarithm
        Y_log = Y.apply(lambda x: np.log(x))
        Y_log = Y_log.drop(labels=['Classic'])
        X = X.drop(['Classic'])
        X = pd.get_dummies(data=X, drop_first=drop_first)
        Y = Y_log
    if preprocessing :
        scaler = MinMaxScaler()
        X_tr=scaler.fit_transform(X)
        X=pd.DataFrame(X_tr, index=X.index, columns=X.columns)
        
    X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.3, random_state=101)
    model = LinearRegression()
    model.fit(X_train,y_train)
    coeff_parameter = pd.DataFrame(model.coef_,X.columns,columns=['Coefficient'])
    predictions = model.predict(X_test)
    sns.regplot(y_test,predictions)
    X_train_Sm= sm.add_constant(X_train)
    ls=sm.OLS(y_train,X_train_Sm).fit()
    print(ls.summary())
    if loga:
        #test_score = model.score(X_test, y_test)
        model_score = model.score(X, Y)
        test_score = msr(np.exp(model.predict(X_test)), np.exp(y_test))
        model_score2 = msr(np.exp(model.predict(X)), np.exp(Y_log))

    
    else:
        #test_score = model.score(X_test, y_test)
        model_score = model.score(X, Y)
        test_score = msr(model.predict(X_test),y_test)
        model_score2 = msr(model.predict(X),Y)


    return coeff_parameter,test_score,model_score,model_score2

    
    
    

def model_analysis_no_plot(X,Y,loga=False,drop_first=False,preprocessing = True):
    if loga:
        # We shouldnt have any negative values so we might try the Logarithm
        Y_log = Y.apply(lambda x: np.log(x))
        Y_log = Y_log.drop(labels=['Classic'])
        X = X.drop(['Classic'])
        X = pd.get_dummies(data=X, drop_first=drop_first)
        Y = Y_log
    if preprocessing :
        scaler = MinMaxScaler()
        X_tr=scaler.fit_transform(X)
        X=pd.DataFrame(X_tr, index=X.index, columns=X.columns)
        
    X_train, X_test, y_train, y_test = train_test_split(X, Y, test_size=0.3, random_state=101)
    model = LinearRegression()
    model.fit(X_train,y_train)
    predictions = model.predict(X_test)

    if loga:

        test_score = msr(np.exp(model.predict(X_test)), np.exp(y_test))
        model_score = msr(np.exp(model.predict(X)), np.exp(Y_log))

    
    else:
        test_score = msr(model.predict(X_test),y_test)
        model_score = msr(model.predict(X),Y)


    return test_score,model_score,model

columns=['Weapon Type', 'Wall Penetration','Fire Rate','Magazine Capacity','Spread ADS','Spread HIP','HDMG_0','BDMG_0','LDMG_0']
X,Y = get_X_y(columns)     
X.head()

	Fire Rate	Magazine Capacity	Spread ADS	Spread HIP	HDMG_0	BDMG_0	LDMG_0	Weapon Type_Heavy	Weapon Type_Rifle	Weapon Type_SMG	Weapon Type_Shotgun	Weapon Type_Sidearm	Weapon Type_Sniper	Wall Penetration_High	Wall Penetration_Low	Wall Penetration_Medium
Name
Classic	6.75	12	0.40	0.40	78	26	22	0	0	0	0	1	0	0	1	0
Shorty	3.33	2	4.00	4.00	24	12	10	0	0	0	0	1	0	0	1	0
Frenzy	13.00	13	0.45	0.45	78	26	22	0	0	0	0	1	0	0	1	0
Ghost	6.75	15	0.30	0.30	105	30	25	0	0	0	0	1	0	0	0	1
Sheriff	4.00	6	0.25	0.25	159	55	46	0	0	0	0	1	0	1	0	0

Let's see some cases

#for preprocessing in True,False :
#    for drop_first in True,False :
#        for loga in  True,False :

preprocessing,drop_first,loga = True, False, True
print('preprocessing:',preprocessing ,"--",'drop_first:',drop_first ,"--",'loga:',loga)
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
#coeff_parameter,test_score,model_score,model_score2
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot,log:{},preprocessing:{},drop:{}".format(loga,preprocessing,drop_first))
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3] )

preprocessing: True -- drop_first: False -- loga: True
                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       1.000
Model:                            OLS   Adj. R-squared:                    nan
Method:                 Least Squares   F-statistic:                       nan
Date:                Sat, 05 Mar 2022   Prob (F-statistic):                nan
Time:                        21:01:37   Log-Likelihood:                 329.24
No. Observations:                  11   AIC:                            -636.5
Df Residuals:                       0   BIC:                            -632.1
Df Model:                          10                                         
Covariance Type:            nonrobust                                         
===========================================================================================
                              coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------------
const                       5.2806        inf          0        nan         nan         nan
Fire Rate                  -0.9532        inf         -0        nan         nan         nan
Magazine Capacity           1.2337        inf          0        nan         nan         nan
Spread ADS                 -2.0044        inf         -0        nan         nan         nan
Spread HIP                  0.3846        inf          0        nan         nan         nan
HDMG_0                      2.0795        inf          0        nan         nan         nan
BDMG_0                     -4.0065        inf         -0        nan         nan         nan
LDMG_0                     -0.1950        inf         -0        nan         nan         nan
Weapon Type_Heavy           0.2138        inf          0        nan         nan         nan
Weapon Type_Rifle           1.3001        inf          0        nan         nan         nan
Weapon Type_SMG             1.2093        inf          0        nan         nan         nan
Weapon Type_Shotgun              0        nan        nan        nan         nan         nan
Weapon Type_Sidearm         0.2272        inf          0        nan         nan         nan
Weapon Type_Sniper          2.3302        inf          0        nan         nan         nan
Wall Penetration_High       2.5440        inf          0        nan         nan         nan
Wall Penetration_Low        1.3701        inf          0        nan         nan         nan
Wall Penetration_Medium     1.3665        inf          0        nan         nan         nan
==============================================================================
Omnibus:                       20.841   Durbin-Watson:                   2.338
Prob(Omnibus):                  0.000   Jarque-Bera (JB):               14.900
Skew:                          -2.161   Prob(JB):                     0.000582
Kurtosis:                       6.718   Cond. No.                         347.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The input rank is higher than the number of observations.
Model test error : 1993498.7978286124
Model total error : 622968.3743214415

The price is positively correlated witht the head damage, but it is weird that's the coefficient is negative for the body damage. The spead is negative as expected, (HIP less infulential), and the fire rate has a negative coeff too. The category is having a considerable effect too.

preprocessing,drop_first,loga = False, True, True
print('preprocessing:',preprocessing ,"--",'drop_first:',drop_first ,"--",'loga:',loga)
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
#coeff_parameter,test_score,model_score,model_score2
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot,log:{},preprocessing:{},drop:{}".format(loga,preprocessing,drop_first))
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3] )

preprocessing: False -- drop_first: True -- loga: True
                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       1.000
Model:                            OLS   Adj. R-squared:                    nan
Method:                 Least Squares   F-statistic:                       nan
Date:                Sat, 05 Mar 2022   Prob (F-statistic):                nan
Time:                        21:01:37   Log-Likelihood:                 313.90
No. Observations:                  11   AIC:                            -605.8
Df Residuals:                       0   BIC:                            -601.4
Df Model:                          10                                         
Covariance Type:            nonrobust                                         
===========================================================================================
                              coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------------
const                       1.4781        inf          0        nan         nan         nan
Fire Rate                   0.1336        inf          0        nan         nan         nan
Magazine Capacity          -0.0020        inf         -0        nan         nan         nan
Spread ADS                  1.6953        inf          0        nan         nan         nan
Spread HIP                 -1.2679        inf         -0        nan         nan         nan
HDMG_0                     -0.0057        inf         -0        nan         nan         nan
BDMG_0                     -0.1538        inf         -0        nan         nan         nan
LDMG_0                      0.3172        inf          0        nan         nan         nan
Weapon Type_Heavy           0.8321        inf          0        nan         nan         nan
Weapon Type_Rifle           0.5062        inf          0        nan         nan         nan
Weapon Type_SMG             0.6632        inf          0        nan         nan         nan
Weapon Type_Shotgun              0        nan        nan        nan         nan         nan
Weapon Type_Sidearm         0.0426        inf          0        nan         nan         nan
Weapon Type_Sniper         -0.5661        inf         -0        nan         nan         nan
Wall Penetration_High       0.2660        inf          0        nan         nan         nan
Wall Penetration_Low        0.1493        inf          0        nan         nan         nan
Wall Penetration_Medium     1.0628        inf          0        nan         nan         nan
==============================================================================
Omnibus:                        0.832   Durbin-Watson:                   0.648
Prob(Omnibus):                  0.660   Jarque-Bera (JB):                0.707
Skew:                           0.490   Prob(JB):                        0.702
Kurtosis:                       2.238   Cond. No.                     4.82e+03
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The input rank is higher than the number of observations.
[3] The condition number is large, 4.82e+03. This might indicate that there are
strong multicollinearity or other numerical problems.
Model test error : 4034693.0969555243
Model total error : 1260841.5927986014

Without scaling, the categorical coeff has more value, which expected (as an equilibrium with the previous case)

preprocessing,drop_first,loga = True, False, False
print('preprocessing:',preprocessing ,"--",'drop_first:',drop_first ,"--",'loga:',loga)
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
#coeff_parameter,test_score,model_score,model_score2
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot,log:{},preprocessing:{},drop:{}".format(loga,preprocessing,drop_first))
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3])

preprocessing: True -- drop_first: False -- loga: False
                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       1.000
Model:                            OLS   Adj. R-squared:                    nan
Method:                 Least Squares   F-statistic:                       nan
Date:                Sat, 05 Mar 2022   Prob (F-statistic):                nan
Time:                        21:01:38   Log-Likelihood:                 272.47
No. Observations:                  11   AIC:                            -522.9
Df Residuals:                       0   BIC:                            -518.6
Df Model:                          10                                         
Covariance Type:            nonrobust                                         
===========================================================================================
                              coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------------
const                    -434.6740        inf         -0        nan         nan         nan
Fire Rate                1382.6125        inf          0        nan         nan         nan
Magazine Capacity        -185.6431        inf         -0        nan         nan         nan
Spread ADS                806.5477        inf          0        nan         nan         nan
Spread HIP               3563.2003        inf          0        nan         nan         nan
HDMG_0                   6051.7006        inf          0        nan         nan         nan
BDMG_0                  -4724.2348        inf         -0        nan         nan         nan
LDMG_0                  -1500.4482        inf         -0        nan         nan         nan
Weapon Type_Heavy         205.0651        inf          0        nan         nan         nan
Weapon Type_Rifle         573.9149        inf          0        nan         nan         nan
Weapon Type_SMG           339.4114        inf          0        nan         nan         nan
Weapon Type_Shotgun              0        nan        nan        nan         nan         nan
Weapon Type_Sidearm     -2049.6501        inf         -0        nan         nan         nan
Weapon Type_Sniper        496.5848        inf          0        nan         nan         nan
Wall Penetration_High    1253.5542        inf          0        nan         nan         nan
Wall Penetration_Low    -1253.3397        inf         -0        nan         nan         nan
Wall Penetration_Medium  -434.8885        inf         -0        nan         nan         nan
==============================================================================
Omnibus:                        1.666   Durbin-Watson:                   0.846
Prob(Omnibus):                  0.435   Jarque-Bera (JB):                1.096
Skew:                          -0.716   Prob(JB):                        0.578
Kurtosis:                       2.416   Cond. No.                         239.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The input rank is higher than the number of observations.
Model test error : 2331658.5946693993
Model total error : 822938.3275303763

Now without the categories

columns=['Fire Rate','Magazine Capacity','Spread ADS','Spread HIP','HDMG_0','BDMG_0','LDMG_0']
X,Y = get_X_y(columns)     
X.head()

	Fire Rate	Magazine Capacity	Spread ADS	Spread HIP	HDMG_0	BDMG_0	LDMG_0
Name
Classic	6.75	12	0.40	0.40	78	26	22
Shorty	3.33	2	4.00	4.00	24	12	10
Frenzy	13.00	13	0.45	0.45	78	26	22
Ghost	6.75	15	0.30	0.30	105	30	25
Sheriff	4.00	6	0.25	0.25	159	55	46

preprocessing,drop_first,loga = True, True, True
#the drop_first doesnt matter here
print('preprocessing:',preprocessing ,"--",'drop_first:',drop_first ,"--",'loga:',loga)
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
#coeff_parameter,test_score,model_score,model_score2
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot,log:{},preprocessing:{},drop:{}".format(loga,preprocessing,drop_first))
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3])

preprocessing: True -- drop_first: True -- loga: True
                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       0.950
Model:                            OLS   Adj. R-squared:                  0.833
Method:                 Least Squares   F-statistic:                     8.105
Date:                Sat, 05 Mar 2022   Prob (F-statistic):             0.0567
Time:                        21:01:38   Log-Likelihood:                 1.3805
No. Observations:                  11   AIC:                             13.24
Df Residuals:                       3   BIC:                             16.42
Df Model:                           7                                         
Covariance Type:            nonrobust                                         
=====================================================================================
                        coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------
const                 7.3929      2.646      2.794      0.068      -1.029      15.815
Fire Rate            -0.6564      2.074     -0.317      0.772      -7.257       5.944
Magazine Capacity     2.9882      1.437      2.080      0.129      -1.584       7.560
Spread ADS           -9.3208      7.878     -1.183      0.322     -34.392      15.750
Spread HIP            8.8874      8.215      1.082      0.359     -17.256      35.031
HDMG_0                6.1060      4.053      1.506      0.229      -6.793      19.005
BDMG_0               15.4863     23.029      0.672      0.549     -57.801      88.773
LDMG_0              -29.5039     34.773     -0.848      0.459    -140.168      81.160
==============================================================================
Omnibus:                        1.388   Durbin-Watson:                   2.748
Prob(Omnibus):                  0.499   Jarque-Bera (JB):                0.081
Skew:                          -0.127   Prob(JB):                        0.960
Kurtosis:                       3.337   Cond. No.                         467.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
Model test error : 979903.678594393
Model total error : 375070.11838569085

It's weird that the Leg damage has a relatively big coefficient (in the absolute value), and that the HIP has a positive coefficient. Note that while the error is big the performance of the model is better than without the spread variables.

preprocessing,drop_first,loga = False, True, False
#the drop_first doesnt matter here
print('preprocessing:',preprocessing ,"--",'drop_first:',drop_first ,"--",'loga:',loga)
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
#coeff_parameter,test_score,model_score,model_score2
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot,log:{},preprocessing:{},drop:{}".format(loga,preprocessing,drop_first))
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3])

preprocessing: False -- drop_first: True -- loga: False
                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       0.985
Model:                            OLS   Adj. R-squared:                  0.949
Method:                 Least Squares   F-statistic:                     27.43
Date:                Sat, 05 Mar 2022   Prob (F-statistic):             0.0101
Time:                        21:01:38   Log-Likelihood:                -71.227
No. Observations:                  11   AIC:                             158.5
Df Residuals:                       3   BIC:                             161.6
Df Model:                           7                                         
Covariance Type:            nonrobust                                         
=====================================================================================
                        coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------
const              -885.7179   1784.363     -0.496      0.654   -6564.359    4792.923
Fire Rate            39.3979     79.660      0.495      0.655    -214.115     292.911
Magazine Capacity    28.1403      4.700      5.987      0.009      13.183      43.098
Spread ADS        -1013.4148    532.605     -1.903      0.153   -2708.400     681.571
Spread HIP         1210.3934    245.350      4.933      0.016     429.581    1991.206
HDMG_0               29.8647      5.852      5.103      0.015      11.240      48.489
BDMG_0              -18.3923     44.543     -0.413      0.707    -160.149     123.364
LDMG_0              -45.5457     51.049     -0.892      0.438    -208.007     116.916
==============================================================================
Omnibus:                        2.506   Durbin-Watson:                   2.201
Prob(Omnibus):                  0.286   Jarque-Bera (JB):                0.330
Skew:                           0.021   Prob(JB):                        0.848
Kurtosis:                       3.848   Cond. No.                     3.56e+03
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
[2] The condition number is large, 3.56e+03. This might indicate that there are
strong multicollinearity or other numerical problems.
Model test error : 1013268.5170922922
Model total error : 373573.136709475

#comparing errors
error_total = dict()
error_test =  dict()
for preprocessing in True,False :
    for drop_first in True,False :
        for loga in  True,False :
            for category in  True,False :
                if category == True :
                    columns = ['Weapon Type', 'Wall Penetration','Fire Rate','Magazine Capacity','HDMG_0','BDMG_0','LDMG_0']
                else :
                    columns = ['Fire Rate','Magazine Capacity','HDMG_0','BDMG_0','LDMG_0']
                X,Y = get_X_y(columns) 
                results = model_analysis_no_plot(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
                error_total[str((preprocessing,drop_first,loga,category))]=results[1]
                error_test[str((preprocessing,drop_first,loga,category))]=results[0]



#plt.bar(list(error_total.keys()), list(error_total.values()))
sns.barplot(x=list(error_total.keys()),y=list(error_total.values()));
plt.xticks(rotation=90)
plt.tight_layout()

sns.barplot(x=list(error_test.keys()),y=list(error_test.values()));
plt.xticks(rotation=90)
plt.tight_layout()

# without the four outliers
outliers= [(True, True, False, True),(True, False, False, True),(False, True, False, True),(False, False, False, True)]
for outlier in outliers:
    del error_total[str(outlier)]
    del error_test[str(outlier)]

sns.barplot(x=list(error_total.keys()),y=list(error_total.values()));
plt.xticks(rotation=90)
plt.tight_layout()

sns.barplot(x=list(error_test.keys()),y=list(error_test.values()));
plt.xticks(rotation=90)
plt.tight_layout()

All the errors are huge, this is not something to model with linear regression , also other variables aren't taken into consideration , like the spray.

#The T,F,T,F example
columns=['Fire Rate','Magazine Capacity','HDMG_0','BDMG_0','LDMG_0']
X,Y = get_X_y(columns)     

preprocessing,drop_first,loga = True, False, True
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot")
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3] )
model =  model_analysis_no_plot(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)[2]

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       0.925
Model:                            OLS   Adj. R-squared:                  0.851
Method:                 Least Squares   F-statistic:                     12.41
Date:                Sat, 05 Mar 2022   Prob (F-statistic):            0.00760
Time:                        21:01:40   Log-Likelihood:               -0.79261
No. Observations:                  11   AIC:                             13.59
Df Residuals:                       5   BIC:                             15.97
Df Model:                           5                                         
Covariance Type:            nonrobust                                         
=====================================================================================
                        coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------
const                 4.8563      0.457     10.637      0.000       3.683       6.030
Fire Rate             1.0460      0.597      1.751      0.140      -0.490       2.582
Magazine Capacity     1.4926      0.506      2.948      0.032       0.191       2.794
HDMG_0                3.5055      0.871      4.025      0.010       1.267       5.744
BDMG_0               -4.4123      9.737     -0.453      0.669     -29.443      20.618
LDMG_0                4.4560      9.532      0.467      0.660     -20.046      28.958
==============================================================================
Omnibus:                        3.527   Durbin-Watson:                   2.492
Prob(Omnibus):                  0.171   Jarque-Bera (JB):                1.004
Skew:                          -0.655   Prob(JB):                        0.605
Kurtosis:                       3.690   Cond. No.                         154.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
Model test error : 1273102.7250278767
Model total error : 486564.3119100869

scaler = MinMaxScaler()
X_tr=scaler.fit_transform(X[1:])
X=pd.DataFrame(X_tr, index=X[1:].index, columns=X[1:].columns)
predicted = pd.Series(np.exp(model.predict(X)), index=Y[1:].index)
predicted = predicted.astype(int)
comparaison = predicted.to_frame(name="predicted")
comparaison["Actual value"] = Y[1:]
comparaison

	predicted	Actual value
Name
Shorty	154	150
Frenzy	832	450
Ghost	839	500
Sheriff	1451	800
Stinger	930	950
Spectre	1102	1600
Bulldog	1401	2050
Guardian	2354	2250
Phantom	3167	2900
Vandal	2889	2900
Marshall	2583	950
Operator	4680	4700
Bucky	197	850
Judge	197	1850
Ares	1326	1600
Odin	3868	3200

#The T,F,T,F example with the spread
columns=['Fire Rate','Magazine Capacity','Spread ADS','Spread HIP','HDMG_0','BDMG_0','LDMG_0']
X,Y = get_X_y(columns)     

preprocessing,drop_first,loga = True, False, True
results = model_analysis(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)
coeff_parameter = results[0]
coeff_parameter.plot(kind="barh", figsize=(9, 7))
plt.title("Coefficients plot")
plt.axvline(x=0, color=".5")
plt.subplots_adjust(left=0.3)
print("Model test error :", results[1])
print("Model total error :",results[3] )
model =  model_analysis_no_plot(X,Y,loga=loga,drop_first=drop_first,preprocessing = preprocessing)[2]

                            OLS Regression Results                            
==============================================================================
Dep. Variable:                  Price   R-squared:                       0.950
Model:                            OLS   Adj. R-squared:                  0.833
Method:                 Least Squares   F-statistic:                     8.105
Date:                Sat, 05 Mar 2022   Prob (F-statistic):             0.0567
Time:                        21:01:40   Log-Likelihood:                 1.3805
No. Observations:                  11   AIC:                             13.24
Df Residuals:                       3   BIC:                             16.42
Df Model:                           7                                         
Covariance Type:            nonrobust                                         
=====================================================================================
                        coef    std err          t      P>|t|      [0.025      0.975]
-------------------------------------------------------------------------------------
const                 7.3929      2.646      2.794      0.068      -1.029      15.815
Fire Rate            -0.6564      2.074     -0.317      0.772      -7.257       5.944
Magazine Capacity     2.9882      1.437      2.080      0.129      -1.584       7.560
Spread ADS           -9.3208      7.878     -1.183      0.322     -34.392      15.750
Spread HIP            8.8874      8.215      1.082      0.359     -17.256      35.031
HDMG_0                6.1060      4.053      1.506      0.229      -6.793      19.005
BDMG_0               15.4863     23.029      0.672      0.549     -57.801      88.773
LDMG_0              -29.5039     34.773     -0.848      0.459    -140.168      81.160
==============================================================================
Omnibus:                        1.388   Durbin-Watson:                   2.748
Prob(Omnibus):                  0.499   Jarque-Bera (JB):                0.081
Skew:                          -0.127   Prob(JB):                        0.960
Kurtosis:                       3.337   Cond. No.                         467.
==============================================================================

Notes:
[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.
Model test error : 979903.678594393
Model total error : 375070.11838569085

scaler = MinMaxScaler()
X_tr=scaler.fit_transform(X[1:])
X=pd.DataFrame(X_tr, index=X[1:].index, columns=X[1:].columns)
predicted = pd.Series(np.exp(model.predict(X)), index=Y[1:].index)
predicted = predicted.astype(int)
comparaison = predicted.to_frame(name="predicted")
comparaison["Actual value"] = Y[1:]
comparaison

	predicted	Actual value
Name
Shorty	152	150
Frenzy	710	450
Ghost	1523	500
Sheriff	329	800
Stinger	828	950
Spectre	1712	1600
Bulldog	1362	2050
Guardian	2380	2250
Phantom	3224	2900
Vandal	2725	2900
Marshall	38	950
Operator	4723	4700
Bucky	127	850
Judge	340	1850
Ares	1247	1600
Odin	3705	3200

scaler.get_params()

{'clip': False, 'copy': True, 'feature_range': (0, 1)}

equation = ['Price'+'=']+['exp(']+[str(round(model.coef_[i], 2))+'*'+'scaler('+columns[i]+')'+'+' for i in range(len(columns))]+[str(model.intercept_)+')']

print(''.join(equation))
            

Price=exp(-0.66*scaler(Fire Rate)+2.99*scaler(Magazine Capacity)+-9.32*scaler(Spread ADS)+8.89*scaler(Spread HIP)+6.11*scaler(HDMG_0)+15.49*scaler(BDMG_0)+-29.5*scaler(LDMG_0)+7.3929436707212215)

warnings.filterwarnings('default')

Matches :

The dataset here is taken from here on kaggle and it's taken from vlr.gg .

Content :

There is four tables. The top level is Matches that will tell you teams playing and match (map) score. Game is the next level that breaks down the specific maps played. Then GameRounds gives a round by round breakdown which shows who won, economy of each team, win type, and buy type, whenever the info is available. The game rounds are packaged in one string that you should be able to cast as a json. Lastly there is GameScoreboard which gives you the player performance, as well as things like number of first kills, first deaths, 2Ks, 3Ks, One v Ones, One v Twos, ect.

This content introduction is made by Joshua Broas on kaagle.

warnings.filterwarnings('ignore')

import sqlite3
import pandas as pd
import sqlalchemy

con = sqlite3.connect(r"C:\Users\anass\Programmation\EDA\Valorant\valorant.sqlite")

cursor = con.cursor()

cursor.execute("SELECT name FROM sqlite_master WHERE type='table';")
tables_names=[name[0] for name in cursor.fetchall()]
cursor.close()
con.close()

cnx = sqlite3.connect(r"C:\Users\anass\Programmation\EDA\Valorant\valorant.sqlite")
list_of_dataframes = []


for name in tables_names:
    list_of_dataframes.append(pd.read_sql_query("SELECT * FROM {}".format(name), cnx))

tables_names

['Matches', 'Games', 'Game_Rounds', 'Game_Scoreboard']

df_matches ,df_games , df_rounds , df_scoreboard =tuple(list_of_dataframes)

del list_of_dataframes

df_matches.head()

	MatchID	Date	Patch	EventID	EventName	EventStage	Team1ID	Team2ID	Team1	Team2	Team1_MapScore	Team2_MapScore
0	62393	2022-01-08 15:30:00	Patch 3.12	826	Nerd Street Gamers Winter Championship - Regio...	Group Stage: Decider (A)	6903	6020	Booster Seat Gaming	Pho Real	2	1
1	62403	2022-01-08 15:30:00	Patch 3.12	826	Nerd Street Gamers Winter Championship - Regio...	Group Stage: Decider (C)	7046	7047	Bjor's Kittens	Mugiwara	2	0
2	62391	2022-01-08 12:30:00	Patch 3.12	826	Nerd Street Gamers Winter Championship - Regio...	Group Stage: Winner's (A)	6461	6903	Akrew	Booster Seat Gaming	2	1
3	62396	2022-01-08 12:30:00	Patch 3.12	826	Nerd Street Gamers Winter Championship - Regio...	Group Stage: Winner's (B)	6164	7043	Radiance	sameROFLMAO	2	0
4	62401	2022-01-08 12:30:00	Patch 3.12	826	Nerd Street Gamers Winter Championship - Regio...	Group Stage: Winner's (C)	7045	7046	Salt and Vinegar	Bjor's Kittens	2	0

df_games.head()

	GameID	MatchID	Map	Team1ID	Team2ID	Team1	Team2	Winner	Team1_TotalRounds	Team2_TotalRounds	...	Team1_FullBuyWon	Team2_PistolWon	Team2_Eco	Team2_EcoWon	Team2_SemiEco	Team2_SemiEcoWon	Team2_SemiBuy	Team2_SemiBuyWon	Team2_FullBuy	Team2_FullBuyWon
0	60894	62393	Breeze	6903	6020	Booster Seat Gaming	Pho Real	Booster Seat Gaming	13	7	...	8.0	0.0	4.0	0.0	2.0	0.0	4.0	1.0	10.0	6.0
1	60895	62393	Bind	6903	6020	Booster Seat Gaming	Pho Real	Pho Real	2	13	...	1.0	2.0	2.0	2.0	0.0	0.0	4.0	3.0	9.0	8.0
2	60896	62393	Haven	6903	6020	Booster Seat Gaming	Pho Real	Booster Seat Gaming	13	8	...	9.0	1.0	2.0	1.0	2.0	0.0	6.0	2.0	11.0	5.0
3	60924	62403	Icebox	7046	7047	Bjor's Kittens	Mugiwara	Bjor's Kittens	13	6	...	8.0	0.0	4.0	0.0	1.0	0.0	2.0	1.0	12.0	5.0
4	60925	62403	Haven	7046	7047	Bjor's Kittens	Mugiwara	Bjor's Kittens	13	9	...	11.0	1.0	3.0	2.0	3.0	0.0	4.0	3.0	12.0	4.0

5 rows × 36 columns

df_rounds.head()

	GameID	Team1ID	Team2ID	RoundHistory
0	60894	6903	6020	{1: {'RoundWinner': 'BOOS', 'ScoreAfterRound':...
1	60895	6903	6020	{1: {'RoundWinner': 'PHO ', 'ScoreAfterRound':...
2	60896	6903	6020	{1: {'RoundWinner': 'PHO ', 'ScoreAfterRound':...
3	60924	7046	7047	{1: {'RoundWinner': 'BJOR', 'ScoreAfterRound':...
4	60925	7046	7047	{1: {'RoundWinner': 'BJOR', 'ScoreAfterRound':...

df_scoreboard.head()

	GameID	PlayerID	PlayerName	TeamAbbreviation	Agent	ACS	Kills	Deaths	Assists	PlusMinus	...	Num_4Ks	Num_5Ks	OnevOne	OnevTwo	OnevThree	OnevFour	OnevFive	Econ	Plants	Defuses
0	60894	8419	Reduxx	Boos	jett	313.0	24.0	10.0	3.0	14.0	...	2.0	0.0	1.0	0.0	0.0	0.0	0.0	74.0	0.0	0.0
1	60894	466	ChurmZ	Boos	chamber	227.0	16.0	10.0	7.0	6.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	67.0	2.0	0.0
2	60894	3712	diaamond	Boos	sova	226.0	17.0	9.0	8.0	8.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	58.0	3.0	0.0
3	60894	5099	Boltzy	Boos	viper	218.0	17.0	12.0	2.0	5.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	48.0	0.0	0.0
4	60894	3983	Virtyy	Boos	skye	80.0	5.0	13.0	3.0	-8.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	21.0	0.0	0.0

5 rows × 28 columns

df_games.head()

	GameID	MatchID	Map	Team1ID	Team2ID	Team1	Team2	Winner	Team1_TotalRounds	Team2_TotalRounds	...	Team1_FullBuyWon	Team2_PistolWon	Team2_Eco	Team2_EcoWon	Team2_SemiEco	Team2_SemiEcoWon	Team2_SemiBuy	Team2_SemiBuyWon	Team2_FullBuy	Team2_FullBuyWon
0	60894	62393	Breeze	6903	6020	Booster Seat Gaming	Pho Real	Booster Seat Gaming	13	7	...	8.0	0.0	4.0	0.0	2.0	0.0	4.0	1.0	10.0	6.0
1	60895	62393	Bind	6903	6020	Booster Seat Gaming	Pho Real	Pho Real	2	13	...	1.0	2.0	2.0	2.0	0.0	0.0	4.0	3.0	9.0	8.0
2	60896	62393	Haven	6903	6020	Booster Seat Gaming	Pho Real	Booster Seat Gaming	13	8	...	9.0	1.0	2.0	1.0	2.0	0.0	6.0	2.0	11.0	5.0
3	60924	62403	Icebox	7046	7047	Bjor's Kittens	Mugiwara	Bjor's Kittens	13	6	...	8.0	0.0	4.0	0.0	1.0	0.0	2.0	1.0	12.0	5.0
4	60925	62403	Haven	7046	7047	Bjor's Kittens	Mugiwara	Bjor's Kittens	13	9	...	11.0	1.0	3.0	2.0	3.0	0.0	4.0	3.0	12.0	4.0

5 rows × 36 columns

df_games.info()

<class 'pandas.core.frame.DataFrame'>
RangeIndex: 15888 entries, 0 to 15887
Data columns (total 36 columns):
 #   Column                   Non-Null Count  Dtype  
---  ------                   --------------  -----  
 0   GameID                   15888 non-null  object 
 1   MatchID                  15888 non-null  object 
 2   Map                      15888 non-null  object 
 3   Team1ID                  15888 non-null  int64  
 4   Team2ID                  15888 non-null  int64  
 5   Team1                    15888 non-null  object 
 6   Team2                    15888 non-null  object 
 7   Winner                   15888 non-null  object 
 8   Team1_TotalRounds        15888 non-null  int64  
 9   Team2_TotalRounds        15888 non-null  int64  
 10  Team1_SideFirstHalf      15888 non-null  object 
 11  Team2_SideFirstHalf      15888 non-null  object 
 12  Team1_RoundsFirstHalf    15888 non-null  int64  
 13  Team1_RoundsSecondtHalf  15888 non-null  int64  
 14  Team1_RoundsOT           15888 non-null  int64  
 15  Team2_RoundsFirstHalf    15888 non-null  int64  
 16  Team2_RoundsSecondtHalf  15888 non-null  int64  
 17  Team2_RoundsOT           15888 non-null  int64  
 18  Team1_PistolWon          14854 non-null  float64
 19  Team1_Eco                14854 non-null  float64
 20  Team1_EcoWon             14854 non-null  float64
 21  Team1_SemiEco            14854 non-null  float64
 22  Team1_SemiEcoWon         14854 non-null  float64
 23  Team1_SemiBuy            14854 non-null  float64
 24  Team1_SemiBuyWon         14854 non-null  float64
 25  Team1_FullBuy            14854 non-null  float64
 26  Team1_FullBuyWon         14854 non-null  float64
 27  Team2_PistolWon          14854 non-null  float64
 28  Team2_Eco                14854 non-null  float64
 29  Team2_EcoWon             14854 non-null  float64
 30  Team2_SemiEco            14854 non-null  float64
 31  Team2_SemiEcoWon         14854 non-null  float64
 32  Team2_SemiBuy            14854 non-null  float64
 33  Team2_SemiBuyWon         14854 non-null  float64
 34  Team2_FullBuy            14854 non-null  float64
 35  Team2_FullBuyWon         14854 non-null  float64
dtypes: float64(18), int64(10), object(8)
memory usage: 4.4+ MB

df_matches.isnull().sum().sum()

475

display(df_scoreboard.describe().round(2))

	ACS	Kills	Deaths	Assists	PlusMinus	KAST_Percent	ADR	HS_Percent	FirstKills	FirstDeaths	...	Num_4Ks	Num_5Ks	OnevOne	OnevTwo	OnevThree	OnevFour	OnevFive	Econ	Plants	Defuses
count	157409.00	157449.00	157449.00	157449.00	156186.00	3367.00	149064.00	148467.00	157409.00	148474.00	...	147744.00	147744.00	147744.00	147744.00	147744.00	147744.00	147744.00	147744.00	147744.00	147744.00
mean	201.13	14.37	14.38	5.15	-0.01	0.70	130.69	0.24	2.03	2.05	...	0.17	0.02	0.20	0.10	0.03	0.00	0.00	53.76	1.36	0.41
std	65.09	5.62	4.06	3.11	6.33	0.13	39.67	0.09	1.74	1.61	...	0.42	0.14	0.45	0.32	0.16	0.07	0.02	18.90	1.66	0.66
min	0.00	0.00	0.00	0.00	-20.00	0.14	0.00	0.00	0.00	0.00	...	0.00	0.00	0.00	0.00	0.00	0.00	0.00	-54.00	0.00	0.00
25%	158.00	10.00	12.00	3.00	-5.00	0.62	103.00	0.17	1.00	1.00	...	0.00	0.00	0.00	0.00	0.00	0.00	0.00	41.00	0.00	0.00
50%	197.00	14.00	15.00	5.00	0.00	0.71	128.00	0.23	2.00	2.00	...	0.00	0.00	0.00	0.00	0.00	0.00	0.00	51.00	1.00	0.00
75%	241.00	18.00	17.00	7.00	4.00	0.79	155.00	0.29	3.00	3.00	...	0.00	0.00	0.00	0.00	0.00	0.00	0.00	64.00	2.00	1.00
max	637.00	56.00	38.00	74.00	30.00	1.00	405.00	1.00	17.00	13.00	...	6.00	2.00	5.00	4.00	3.00	1.00	1.00	566.00	15.00	6.00

8 rows × 23 columns

warnings.filterwarnings('ignore')

sns_plot = sns.distplot(df_scoreboard["Kills"])

sns_plot = sns.distplot(df_scoreboard["Assists"])

sns_plot = sns.distplot(df_scoreboard["Plants"])

df_scoreboard.head()

	GameID	PlayerID	PlayerName	TeamAbbreviation	Agent	ACS	Kills	Deaths	Assists	PlusMinus	...	Num_4Ks	Num_5Ks	OnevOne	OnevTwo	OnevThree	OnevFour	OnevFive	Econ	Plants	Defuses
0	60894	8419	Reduxx	Boos	jett	313.0	24.0	10.0	3.0	14.0	...	2.0	0.0	1.0	0.0	0.0	0.0	0.0	74.0	0.0	0.0
1	60894	466	ChurmZ	Boos	chamber	227.0	16.0	10.0	7.0	6.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	67.0	2.0	0.0
2	60894	3712	diaamond	Boos	sova	226.0	17.0	9.0	8.0	8.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	58.0	3.0	0.0
3	60894	5099	Boltzy	Boos	viper	218.0	17.0	12.0	2.0	5.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	48.0	0.0	0.0
4	60894	3983	Virtyy	Boos	skye	80.0	5.0	13.0	3.0	-8.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	21.0	0.0	0.0

5 rows × 28 columns

fig, ax = plt.subplots(5, 2, figsize = (15, 13))
sns.boxplot(x= df_scoreboard["Kills"], ax = ax[0,0])
sns.distplot(df_scoreboard['Kills'], ax = ax[0,1])
sns.boxplot(x= df_scoreboard["Deaths"], ax = ax[1,0])
sns.distplot(df_scoreboard['Deaths'], ax = ax[1,1])
sns.boxplot(x= df_scoreboard["Assists"], ax = ax[2,0])
sns.distplot(df_scoreboard['Assists'], ax = ax[2,1])
sns.boxplot(x= df_scoreboard["Plants"], ax = ax[3,0])
sns.distplot(df_scoreboard['Plants'], ax = ax[3,1])
sns.boxplot(x= df_scoreboard["Defuses"], ax = ax[4,0])
sns.distplot(df_scoreboard['Defuses'], ax = ax[4,1])
plt.tight_layout()

df_scoreboard_cleaned = df_scoreboard.dropna()
df_scoreboard_cleaned

	GameID	PlayerID	PlayerName	TeamAbbreviation	Agent	ACS	Kills	Deaths	Assists	PlusMinus	...	Num_4Ks	Num_5Ks	OnevOne	OnevTwo	OnevThree	OnevFour	OnevFive	Econ	Plants	Defuses
0	60894	8419	Reduxx	Boos	jett	313.0	24.0	10.0	3.0	14.0	...	2.0	0.0	1.0	0.0	0.0	0.0	0.0	74.0	0.0	0.0
1	60894	466	ChurmZ	Boos	chamber	227.0	16.0	10.0	7.0	6.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	67.0	2.0	0.0
2	60894	3712	diaamond	Boos	sova	226.0	17.0	9.0	8.0	8.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	58.0	3.0	0.0
3	60894	5099	Boltzy	Boos	viper	218.0	17.0	12.0	2.0	5.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	48.0	0.0	0.0
4	60894	3983	Virtyy	Boos	skye	80.0	5.0	13.0	3.0	-8.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	21.0	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
21883	53281	2126	Shawn	GEN	sage	196.0	12.0	13.0	3.0	-1.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	43.0	3.0	0.0
21884	53281	4927	NaturE	GEN	jett	149.0	10.0	13.0	1.0	-3.0	...	1.0	0.0	0.0	0.0	0.0	0.0	0.0	39.0	0.0	1.0
21885	53281	156	Temperature	GEN	sova	123.0	7.0	12.0	3.0	-5.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	44.0	0.0	1.0
21886	53281	64	gMd	GEN	omen	121.0	6.0	16.0	4.0	-10.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	32.0	0.0	0.0
21887	53281	8716	koosta	GEN	viper	101.0	5.0	14.0	7.0	-9.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	35.0	0.0	0.0

3367 rows × 28 columns

#create scatterplot of hours vs. score
plt.scatter(df_scoreboard["Kills"], df_scoreboard["Deaths"])
plt.title('Hours Studied vs. Exam Score')
plt.xlabel('Hours Studied')
plt.ylabel('Exam Score')

Text(0, 0.5, 'Exam Score')

#create scatterplot of hours vs. score
plt.scatter(df_scoreboard["Kills"], df_scoreboard["Assists"])
plt.title('Kills vs. Assists')
plt.xlabel('Kills')
plt.ylabel('Assists')

Text(0, 0.5, 'Assists')

from fitter import Fitter, get_common_distributions, get_distributions

Kills = df_scoreboard["Kills"].values
Kills = Kills[np.logical_not(np.isnan(Kills))] 
f = Fitter(Kills,
           distributions=['gamma',
                          'lognorm',
                          "beta",
                          "burr",
                          "norm"])
f.fit()
f.summary()

	sumsquare_error	aic	bic	kl_div
lognorm	0.071433	1576.168643	-2.299640e+06	inf
beta	0.071440	1600.478699	-2.299614e+06	inf
gamma	0.071440	1598.667604	-2.299626e+06	inf
norm	0.071851	1947.547679	-2.298734e+06	inf
burr	0.073985	1130.418124	-2.294103e+06	inf

from distfit import distfit

# Initialize model
dist = distfit()

# Find best theoretical distribution for empirical data X
dist.fit_transform(Kills)
dist.plot()

[distfit] >fit..
[distfit] >transform..
[distfit] >[norm      ] [0.00 sec] [RSS: 0.0039462] [loc=14.368 scale=5.623]
[distfit] >[expon     ] [0.00 sec] [RSS: 0.0309937] [loc=0.000 scale=14.368]
[distfit] >[pareto    ] [2.15 sec] [RSS: 0.0644799] [loc=-2.696 scale=2.696]
[distfit] >[dweibull  ] [1.19 sec] [RSS: 0.0041518] [loc=13.608 scale=4.840]
[distfit] >[t         ] [2.80 sec] [RSS: 0.0040067] [loc=14.322 scale=5.416]
[distfit] >[genextreme] [5.77 sec] [RSS: 0.0039802] [loc=11.953 scale=5.301]
[distfit] >[gamma     ] [1.04 sec] [RSS: 0.0037878] [loc=-31.107 scale=0.694]
[distfit] >[lognorm   ] [5.98 sec] [RSS: 0.0037917] [loc=-52.337 scale=66.470]
[distfit] >[beta      ] [3.61 sec] [RSS: 0.0037878] [loc=-31.072 scale=11023959.598]
[distfit] >[uniform   ] [0.00 sec] [RSS: 0.0332003] [loc=0.000 scale=56.000]
[distfit] >[loggamma  ] [2.24 sec] [RSS: 0.0039749] [loc=-1368.749 scale=195.088]
[distfit] >Compute confidence interval [parametric]
[distfit] >plot..

(<Figure size 720x576 with 1 Axes>,
 <AxesSubplot:title={'center':'\nbeta\na=65.38, b=15860599.41, loc=-31.07, scale=11023959.60'}, xlabel='Values', ylabel='Frequency'>)

Agents :

df_scoreboard

	GameID	PlayerID	PlayerName	TeamAbbreviation	Agent	ACS	Kills	Deaths	Assists	PlusMinus	...	Num_4Ks	Num_5Ks	OnevOne	OnevTwo	OnevThree	OnevFour	OnevFive	Econ	Plants	Defuses
0	60894	8419	Reduxx	Boos	jett	313.0	24.0	10.0	3.0	14.0	...	2.0	0.0	1.0	0.0	0.0	0.0	0.0	74.0	0.0	0.0
1	60894	466	ChurmZ	Boos	chamber	227.0	16.0	10.0	7.0	6.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	67.0	2.0	0.0
2	60894	3712	diaamond	Boos	sova	226.0	17.0	9.0	8.0	8.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	58.0	3.0	0.0
3	60894	5099	Boltzy	Boos	viper	218.0	17.0	12.0	2.0	5.0	...	0.0	0.0	1.0	0.0	0.0	0.0	0.0	48.0	0.0	0.0
4	60894	3983	Virtyy	Boos	skye	80.0	5.0	13.0	3.0	-8.0	...	0.0	0.0	0.0	0.0	0.0	0.0	0.0	21.0	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
157934	13	24	Gover			0.0	0.0	0.0	0.0	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
157935	13	25	Jack1			0.0	0.0	0.0	0.0	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
157936	13	26	Rewind			0.0	0.0	0.0	0.0	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
157937	13	27	Woo1y			0.0	0.0	0.0	0.0	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN
157938	13	28	DrasseL			0.0	0.0	0.0	0.0	NaN	...	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN	NaN

157939 rows × 28 columns

len(df_agents)

18

df_agents= df_scoreboard.groupby("Agent").mean().sort_values(by='ACS', ascending=False)
df_agents.head()

	ACS	Kills	Deaths	Assists	PlusMinus	KAST_Percent	ADR	HS_Percent	FirstKills	FirstDeaths	...	Num_4Ks	Num_5Ks	OnevOne	OnevTwo	OnevThree	OnevFour	OnevFive	Econ	Plants	Defuses
Agent
raze	235.210910	16.141252	15.286146	4.715481	0.855106	0.686279	150.846141	0.180850	2.853556	2.775425	...	0.226162	0.029545	0.151125	0.080460	0.016675	0.002914	0.000486	59.121661	0.591145	0.427311
reyna	230.294592	16.426312	15.196395	4.201929	1.229918	0.690932	146.345062	0.259528	2.904016	2.838500	...	0.274841	0.040169	0.167995	0.081965	0.021142	0.004228	0.000163	58.550171	0.490649	0.287689
jett	230.045706	16.582086	15.028903	3.306768	1.553183	0.682714	141.720387	0.229524	3.656197	3.025411	...	0.262311	0.034978	0.166248	0.082346	0.020850	0.003612	0.000366	58.974258	0.496502	0.344337
phoenix	221.685185	15.572784	15.042814	4.794029	0.529970	0.630000	140.427653	0.252350	2.768750	2.689311	...	0.216443	0.027150	0.156306	0.081198	0.019792	0.003552	0.000000	56.124841	0.747780	0.485156
yoru	213.034483	15.088670	15.487685	4.339901	-0.399015	NaN	135.788177	0.250493	2.743842	2.817734	...	0.246305	0.024631	0.187192	0.064039	0.024631	0.000000	0.000000	52.837438	0.748768	0.285714

5 rows × 23 columns

#df_scoreboard.groupby(by="Agent").sum().sort_values(by='Kills', ascending=False)

def return_sorted2(df_new,col_name):
    sorted_df = df_new.sort_values(by=col_name, ascending=False)
    return {'Agent': sorted_df.index.to_list(), col_name: sorted_df[col_name].to_list()}

ACS_dict = return_sorted2(df_agents,'ACS')

fig_ACS= px.bar(ACS_dict, x = ACS_dict['Agent'], y = 'ACS', title = 'Average number of ACS by Agent')
fig_ACS.show()

kills_dict = return_sorted2(df_agents,'Kills')

fig_kills= px.bar(kills_dict, x = kills_dict['Agent'], y = 'Kills', title = 'Average number of Kills by Agent')
fig_kills.show()

Assists_dict = return_sorted2(df_agents,'Assists')

fig_Assists= px.bar(Assists_dict, x = Assists_dict['Agent'], y = 'Assists', title = 'Average number of Assists by Agent')
fig_Assists.show()

Deaths_dict = return_sorted2(df_agents,'Deaths')

fig_Deaths= px.bar(Deaths_dict, x = Deaths_dict['Agent'], y = 'Deaths', title = 'Average number of Deaths by Agent')
fig_Deaths.show()

FirstDeaths_dict = return_sorted2(df_agents,'FirstDeaths')

fig_FirstDeaths= px.bar(FirstDeaths_dict, x = Deaths_dict['Agent'], y = 'FirstDeaths', title = 'Average number of FirstDeaths by Agent')
fig_FirstDeaths.show()

Plants_dict = return_sorted2(df_agents,'Plants')

fig_Plants= px.bar(Plants_dict, x = Deaths_dict['Agent'], y = 'Plants', title = 'Average number of Plants by Agent')
fig_Plants.show()